Logo
Brightwave

Software Engineer - Search

Brightwave, New York City, NY, United States


?? About Brightwave

Brightwave is a deep technology startup building LLM-powered investment research systems that can autonomously reason across huge bodies of multimodal text, table and image data such as SEC filings, real-time news feeds, earnings call transcripts, legal rulings, patents, sell-side analysis and more. Brightwave helps active asset managers accelerate research workflows and generate investment ideas with unprecedented depth and comprehensiveness. Brightwave is experiencing rapid growth and commercial success, with paying customers with assets under management in excess of $120B and brisk demand from diverse clients across financial services.

??? Create Something Incredible

We hire exceptional humans who are intrinsically motivated to do great work. As a principle we tend to favor a smaller team of senior, seasoned professionals who intuitively know what needs to happen next. Brightwave's team includes engineers from places like Databricks and Meta, NeurIPS-published research scientists, and commercial talents with tours of duty at firms like Goldman Sachs, UBS and McKinsey.

Our venture investors include some of the world's largest asset managers as well as angels from OpenAI, Databricks, Uber, LinkedIn and the Solana Foundation.

Search Engineering at Brightwave

Search engineering at Brightwave is responsible for architecting our multi-index search infrastructure, with a strong focus on search relevance optimization. We view Retrieval-Augmented Generation (RAG) systems primarily through the lens of information retrieval and are looking for an individual with deep expertise in the search and IR domain.

This role will involve retrieving and integrating content from highly heterogeneous corpuses, including SEC filings, earnings call transcripts, breaking news, tabular data, images, social media content, podcast transcripts, patents, and government records. You’ll be responsible for building the systems that ingest, search through and unify these diverse sources of data to create a consistently accurate, useful product experience.

?? Responsibilities

  • Create and implement advanced search relevance algorithms to address specific IR challenges, including embedding-based methods, cross-encoders, full-text search, and hybrid search strategies.
  • Design federated search systems to unify results across multiple document types (e.g., SEC filings, patents, podcasts, tabular data, etc.).
  • Define and deploy domain-specific measurement techniques to evaluate and continuously improve search quality.
  • Manage and operate heterogeneous, multimodal search indices in our various Kubernetes clusters, ensuring high performance and availability.
  • Design and implement the serving infrastructure for GPU-based models.
  • Implement thoughtfully-designed API's that expose the search infrastructure you build to internal and external consumers.

?? Capabilities

Strong candidates for this role will have some (but not necessarily all) of the following capabilities and experiences. At Brightwave, we believe in growing together.

  • 7+ years relevant experience in software engineering, with 3+ specializing in search.
  • M.S. or Ph.D. in computer science, statistics, or other relevant domain.
  • Experience building, fitting and serving classical and transformer-based deep learning models.
  • Exposure to frameworks like pytorch, hugingface, sentence-transformers, sklearn, etc.
  • Experience in training data curation, including human annotation, impression/action tracking, and synthetic data generation.
  • Production experience with database technologies like ElasticSearch, Postgres, Qdrant, FAISS, Neo4J, etc.
  • Understanding techniques for knowledge graph construction, including named entity recognition, entity linking, and edge prediction.
  • Experience with large-scale data processing and data pipeline orchestration systems (e.g., Spark, Dagster, Airflow, etc.)
  • Exposure toRetrieval-Augmented Generation (RAG) systems and their application to diverse datasets.
  • Strong computer science, machine learning, and statistical fundamentals.

Values Matter

We stay true to our values, and work with people who do the same.

  • Ownership. We take responsibility for how we choose to respond to our circumstances.
  • Humility. Most situations warrant a sense of curiosity and a willingness to be wrong.
  • Craftsmanship. We take pride in the work and tend to the details.
  • Focus. Transformative work only happens in the presence of intense, sustained focus.
  • Empathy. Seek first to understand - most people are basically reasonable and want to do the right thing.
  • Humor. Life is a gift, let’s enjoy it together.